Probability of causation for multi-causal occupational disease compensation: A scoping review of methods and alternatives

This is the authors' abstract. We don't add key points for this paper.

Background: Diseases of multi-factorial etiology pose challenges for attributing causation to specific exposures. The probability of causation (PoC) is commonly used to determine whether a worker's cancer is more likely than not due to occupational exposure, but has limitations (e.g., not accounting for acceleration of cancer occurrence). Objectives: To describe PoC methods applied for occupational lung cancer compensation (PoC review), and improvements or alternatives for compensating workers with multi-causal occupational diseases (Alternatives review). Sources of evidence: PubMed and Embase were queried for both reviews (May 2025). The PoC review additionally searched OpenAlex with citation snowballing (n=1,697 records); CINAHL was additionally queried for the Alternatives review (n=329). No language restrictions were applied. Registered protocol: https://osf.io/9evsy. Charting methods: Titles and abstracts were screened using ASReview. Data extracted included chemical/physical agent, PoC thresholds, uncertainty sources, and methodological improvements or alternatives for compensation. Results: Fifteen studies were selected for the PoC review. Radiation was the most assessed agent (n=11), followed by asbestos (n=3), and coal-tar pitch (n=1). Thirteen studies applied a PoC threshold of 50%, using upper 95% (n=8) or 99% (n=2) confidence intervals to favor claimants. Main uncertainty sources: statistical error (n=10), population risk transfer (n=7), and exposure measurement error (n=6). The Alternatives review included 12 studies addressing mechanistic models of cancer development, years of life lost, and frameworks considering harm dimensions beyond binary compensation eligibility. Conclusions: PoC was applied for three of 20 established occupational lung carcinogens. Methodological improvements of PoC estimation and alternative methods could improve individualized compensation assessments.